The present invention relates to the determination of the complexed forms of immunologically determinable prostate specific antigen (PSA) in a blood sample. More particularly, the invention relates to the determination of complexed PSA by two-site immunometric assay and the clinical significance of complexed PSA assay values.
Human prostate specific antigen (PSA) is a glycoprotein of approximately 33,000 daltons with high amino acid homology to the human kallikrein family (1,2) and has been shown to be a serine protease with trypsin and chymotrypsin-like activity (3,4,5). PSA is secreted by epithelial cells of the prostate gland and is one of the major proteins found in seminal fluid (6). Following the discovery that the concentration of PSA increases in the serum of patients with prostate cancer, numerous reports have established this protein as an important and clinically useful biomarker for the management of prostate cancer patients (7,8,9,10). Recent efforts have focused on the use of serum PSA testing for early detection of prostate cancer in asymptomatic men. In fact, the American Cancer Society and the American Urological Society have recently recommended that all men over the age of 50 be screened annually using serum PSA in conjunction with digital rectal examination (DRE) (11).
The clinical value of early detection of prostate cancer remains controversial for several reasons. First, it is unclear whether treatment of prostate cancer at early stages will improve survival in the affected population. Clinical trials designed to address this issue are currently underway. Second, a clinical trial recently measured the effectiveness of serum PSA measurements in conjunction with digital rectal examination (DRE) for early detection of prostate cancer in men over 50 years of age (12). Of the 1060 patients who had either an abnormal DRE or an elevated PSA test, only 22% had prostate cancer. These data demonstrate that 70-80% of all prostate biopsies are performed on men who do not have cancer. Since 30-50% of men over the age of 50 have evidence of prostate cancer on autopsy, the number of unnecessary prostate biopsies triggered by elevated PSA assays could be very high. This has consequences both in increased medical costs and increased morbidity associated with the biopsy procedure.
Several laboratories have shown independently that PSA forms complexes with protease inhibitors such as .alpha..sub.1 -antichymotrypsin (ACT), .alpha..sub.2 -macroglobulin, and .alpha..sub.1 -antitrypsin (13-19). PSA in complex with ACT or .alpha..sub.1 -antitrypsin or in free, uncomplexed form is detectable in serum by immunossay techniques. Indeed, the majority of immunoreactive PSA in serum is complexed with ACT, and a significant correlation has been established between the proportion of PSA bound to ACT and total serum PSA concentration (13). PSA bound to .alpha..sub.2 -macroglobulin, however, is not measurable in serum due to steric hindrance of antibody binding to PSA following complexation with this protease. In early work, PSA-ACT levels and the proportion of PSA-ACT to total PSA were suggested to be of use in prostate cancer diagnosis (13,15,16,17), however, for a variety of reasons (some of which are discussed below) it has been difficult to draw conclusions on the clinical utility of serum measurement of PSA-ACT.
Lilja, Stenman, and coworkers published in 1991 that serum PSA exists in free form and in complexes with ACT and .alpha..sub.1 -antitrypsin (13,18). In subsequent work, Stenman et al. demonstrated that measurement of PSA-ACT in association with measurement of free plus completed PSA (termed total PSA, although PSA completed with .alpha..sub.2 -macroglobulin is not measured by conventional PSA assays) may improve discrimination between men with prostate cancer and those with benign prostate disease such as benign prostatic hypertrophy (BPH). However, the accurate measurement of PSA-ACT complexes has not been attainable due to technical problems in accurate measurement of the complex. Stenman et al. found that the correlation of PSA-ACT values with total PSA measurement was not good at the low end and the y intercept was elevated indicating over-recovery of completed PSA (13 and U.S. Pat. No. 5,501,983). Indeed, they found that for most patients tested, the concentration of PSA-ACT was higher than for total PSA (U.S. Pat. No. 5,501,983). Subsequent correlation analysis for complexed and free PSA-showed a slope of 1.12 indicating over-recovery of the PSA-ACT complex (16). Pettersson et al. addressed this over-recovery when they found elevated PSA-ACT values in female sera (20). While the addition of heparin reduced the incidence of false positive values in female serum, more recent attempts to measure PSA-ACT complexes in patients with prostate cancer and BPH continue to show significant over-recovery of complexes (21).
Because of the difficulties encountered in the measurement of PSA-ACT complexes, attention in the literature turned to the measurement of free, uncomplexed PSA in conjunction with measurement of total PSA. It is now clear that improved specificity is needed when total PSA values range from about 4-10 ng/mL. When serum total PSA is &lt;4.0 ng/mL, the risk of prostate cancer is low; similarly, when total PSA is &gt;10 ng/mL, the risk of prostate cancer is &gt;50% and prostate biopsy is indicated. Within the diagnostic gray zone (generally between 2-20 ng/mL, more commonly between 4-10 ng/mL) the risk of cancer is high, but the rate of false positives is also high. The retrospective application of a ratio of free PSA/total PSA has shown that the specificity of total PSA in the 4-10 ng/mL gray zone could be improved from approximately 50-60% to 70-80% (22-26). This improved specificity could result in a 20-30% decrease in unnecessary biopsies. PCT WO 96-26441 and WO 97-12245 similarly describe the use of the free PSA/total PSA ratio to improve discrimination between BPH and cancer, respectively, in patients with total PSA levels between 2.5 and 20 ng/mL.
The measurement of free PSA has technical difficulties of its own, however. First, within the diagnostic gray zone, the proportion of free PSA is typically quite low, in the 5-30% range. A successful free PSA assay must, therefore, measure accurately in the range of 0.2-3.0 ng/mL. Also, the concentration of free PSA is not significantly different in patients with BPH and cancer, and the ratio of free PSA/total PSA decreases due to an increase in the proportion of PSA complexed to ACT. In addition, free PSA is not stable in serum and levels of free PSA have been known to decrease over time, presumably due to complexation with .alpha..sub.2 -macroglobulin.
In the meantime, there has been further acknowledgment of the problems associated with the accurate measurement of PSA-ACT in blood, coupled with attempts to overcome such problems. In 1994, workers at Hybritech reported the development of a sandwich immunoassay for PSA-ACT employing anti-PSA and anti-ACT antibodies. They concluded that the measured PSA-ACT values failed to demonstrate improved clinical specificity in the diagnosis of prostate cancer (27). Later, this group jointly with workers at the Johns Hopkins Medical Institutions reported the finding that the anti-PSA/anti-ACT sandwich immunoassay method suffers from significant non-specific binding and over recovery of PSA-ACT. Unless resolved, they concluded that these problems rendered the measurement of PSA-ACT clinically meaningless (28). Subsequently, this joint group reported having overcome the non-specific binding problem through the development of a sandwich immunoassay for PSA-ACT based on a monoclonal antibody specific for PSA-ACT complex (29, 30). However, their clinical studies failed to show any improvement in specificity for prostate cancer by measuring PSA-ACT complex alone compared with measurement of total PSA or with a calculated ratio of PSA-ACT to total PSA (29). Other approaches to overcoming the problems associated with PSA-ACT measurement have been proposed, including the use of blocking agents (31).
It remains unclear why the proportion of PSA complexed to ACT increases in patients with prostate cancer, but it may be related to the observation that antibodies to ACT do not stain prostatic epithelium from BPH patients and mRNA transcripts are not found in such tissue. In contrast, anti-ACT immunoreactivity and mRNA synthesis are detected in prostatic epithelium from patients with prostate cancer (32). These results suggest that in prostate tumors, PSA may complex in situ with ACT prior to release into serum. An alternative mechanism may involve the access of active PSA to the blood stream. Free PSA found in serum from healthy men is proteolytically cleaved and enzymatically inactive. Tumors, however, synthesize angiogenic factors which lead to increased vascularization of tumor tissues. It may be that in tumors, a larger proportion of enzymatically active PSA gains access to the blood stream. This active PSA would be expected to complex with protease inhibitors such as ACT leading to a higher proportion of PSA-ACT complex in serum from prostate cancer patients.
Accordingly, there is a need for an accurate method of determining complexed PSA and to assess the clinical significance of blood levels of complexed PSA relative to screening of male patients for prostate cancer.
EP 635,575 describes the preparation of monoclonal antibodies that bind to free PSA but not PSA-ACT.
PCT WO 95/18381 relates to a monoclonal/polyclonal immunometric assay method for the determination of PSA which is rendered capable of providing an equimolar response to free and complexed PSA by the addition of antibody that binds to free PSA but not complexed PSA.
U.S. patent application Ser. No. 08/595,155 now abandoned and Zhou Z, Ng P C, Very D L, Allard W J, Yeung K K, J. Clin. Lab. Anal. (1996), 10:155-159, describe a method for preparing a monoclonal antibody that provides an equimolar response to free and complexed PSA in a monoclonal/polyclonal immunometric assay. The described monoclonal antibody has the unique property of binding to PSA to render PSA substantially incapable of binding with antibodies that bind free PSA but not complexed PSA.
Published Japanese Patent Document 62-46263 describes a sandwich immunoassay method for the determination of PSA in complex with protease inhibitor.
Published German Patent Application 4,322,342 describes a method for measuring both total PSA and PSA-ACT in a single assay for the purpose of providing values for calculation of the ratio of PSA-ACT to total PSA.
Chichibu et al, in the Journal of Medicine and Pharmaceutical Science (Japan, 1996) 36(3): 477-483, describe a sandwich immunoassay for PSA-ACT employing anti-PSA bound to a bead and enzyme-labeled anti-ACT. Data establishing the ability to accurately measure PSA-ACT in a blood sample is lacking.